Development of Entrustable Professional Activities for the University of New Mexico Nephrology Fellowship Training Program

Abstract

Background: Entrustable Professional Activities (EPAs) that transform competencies into distinct, assessable clinical tasks have not yet been developed for US nephrology fellowships. We created and achieved consensus on a set of nephrology-specific EPAs and aligned them with Accreditation Council for Graduate Medical Education (ACGME) competency standards. Methods: This study was conducted within the University of New Mexico nephrology fellowship program. An initial EPA list was generated by the study team using program objectives, a literature review, and clinician insight. Study participants included eight faculty nephrologists and one nephrology fellow, who completed an online-based three-round modified Delphi consensus-building processes. Each EPA was rated on a five-point Likert scale with consensus requiring strict criteria. Finalized EPAs were independently mapped to ACGME nephrology program requirements. Results: Nine study participants (100% response rate) completed all survey rounds. Through iterative consensus, utilizing strict criteria, a final list of 22 distinct EPAs was created, covering 10 core domains of practice including dialysis management, acute kidney injury, chronic kidney disease, electrolyte abnormalities, hypertension, kidney stones, glomerular disease, pregnancy, transplant care, and education. Finalized EPAs were mapped to 38 different ACGME-required sub-competencies, showcasing diversity and applicability to national expectations. Conclusions: We developed the first consensus-based set of EPAs geared for US nephrology fellowship programs, providing a foundation for standardized assessment and curriculum development that could be implemented across nephrology fellowship programs nationally.

1. Introduction

Competency-Based Medical Education (CBME) has transformed postgraduate medical training by shifting the focus from time-based progression to the demonstration of essential skills and competencies required for independent clinical practice [1]. CBME is grounded in clearly defined competencies that trainees must acquire, often structured around frameworks such as the Accreditation Council for Graduate Medical Education (ACGME) competencies and milestones. However, while these frameworks establish broad domains of competency, they often lack direct applicability to day-to-day clinical decision-making and real-world practice [2]. To bridge this gap, Entrustable Professional Activities (EPAs) have emerged as a practical tool within CBME, translating these competencies into specific, observable, and assessable clinical tasks that reflect the actual responsibilities of a physician.

EPAs are formulated units of clinical work that can quantifiably measure the competency of a professional in training. By aligning assessment with workplace-based tasks, EPAs provide a structured approach to evaluating clinical competence, ensuring standardization in training while allowing for individualized learning progression.

When looking at specialties outside of nephrology that have adopted EPA usage for trainee growth, a landmark study in pediatrics found that EPA-based assessments showcased statistically significant growth across their training years over the course of residency [3]. Others emphasize that the implementation of EPAs offers trainees consistent feedback grounded in specific areas based on direct observed performance that may not be available with older evaluation methods [4]. Even when observing the ease of comprehension, over 89% of fellowship program directors in pediatrics across the United States agreed that EPAs were easier to understand than milestones and were a better representation of the traits of a self-sufficient, post-fellowship attending [5].

The main aspect of EPAs includes the performance of a task by a trainee under direct supervision. The passing of responsibility to a trainee to complete a task has been deemed the entrustment decision [6,7]. A supervising physician will entrust a task to a trainee which they deem appropriate for their level of competency. EPAs are uniquely suited for the medical profession as most daily activities of care and treatment that a provider performs can be broken down into self-contained tasks that can be individually tracked. This allows a training program to monitor and assess the competency level of their trainees through quantifiable metrics which they can apply to the broader ACGME competencies.

Currently, the training and competency-based assessment we provide and use in nephrology fellowship programs can leave gaps in content and clinical knowledge necessary for successful independent practice, especially for those graduates entering non-academic practices. This is not to suggest that the current training structures of nephrology fellows, without EPAs, are critically lacking in specific content but rather an opportunity to offer a way to standardize future training. Our graduating fellows must develop skillsets aligned to national practice environments, including patient safety and development of professional practice. EPAs are a tool that programs could utilize to help ensure their fellows are successfully developing these necessary skills.

The most recent study on the development of nephrology-specific EPAs comes from Tanaka et al., who applied a modified Delphi method to inform the development of EPAs for IM residents rotating through a nephrology department in a Japanese academic hospital [8]. While narrower in scope and geared towards a different learner population, Tanaka’s methodology strongly influenced the design of our own study. Specifically, we adopted their strict consensus criteria and the use of modified Delphi method for constructing a consensus list of program-specific EPAs [8]. Even with the existence of Tanaka’s work, there is still a lack of consensus regarding the development of EPAs, their implementation, and their use specifically in American Nephrology fellowship programs.

Therefore, in this study, we sought to address this unmet need to develop EPAs for nephrology fellowship training. These EPAs could be applied after site-specific modifications to other programs as well. We first focused on determining learners’ and educators’ perspectives and needs regarding work-based assessment to develop EPAs for our nephrology training program at the University of New Mexico (UNM). We aimed to develop a list of EPAs that represent the scope of practice in nephrology both nationwide and in the state of New Mexico, NM. In addition, we matched the EPAs we developed to the available ACGME competencies to create concrete goals and milestones for fellows.

2. Materials and Methods

All study forms and procedures were approved by UNM-HSC’s IRB (HRPO #23-338).

2.1. Creation of Initial EPA List

The initial list of EPAs was developed by members of the research group after three online meetings (from October 2023 to April 2024). The group included four experienced attending physicians, one medical education specialist, and two 2nd-year nephrology fellows. The initial list was created by referencing and adapting annually reviewed training goals and behavioral objectives of the Division of Nephrology at UNM. Textbook chapter headings and search lists of the most common diagnostic codes were considered. An iterative process was used to develop this initial list, reorganizing after each of several discussions and adhering to the following conditions [9]:

• The EPAs should be part of the essential professional work of nephrology and not general medical ability.
• Must require adequate knowledge, skill, and attitude.
• Must lead to recognized performance that is unique to a doctor and should be unique to physicians in nephrology.
• They should be independently executable within a specific time frame.
• They should be observable and measurable in their process and outcome.
• They should reflect one or more of the ACGME competency categories.

The development of each EPA’s description followed the guidelines set by ten Cate [10] (Table S1). Based on a review of the literature and the 23 ACGME sub-competencies required for nephrology training, we hypothesized that we would have to develop at least 10 EPAs to be able to cover most of the various clinical aspects of the work of a practicing nephrologist.

2.2. Consensus Development Method

Similar to prior studies which successfully developed EPAs, we used the modified Delphi method as our consensus-building approach [8].

We found that the modified Delphi approach was particularly well-suited to the design of our study that consisted of a pre-constructed list which was iterated in an asynchronous setting, maintaining anonymity amongst our participants.

2.3. Recruitment of Survey Respondents

After a brief presentation of the project to the nephrology division, potential study participants were approached through a confidential e-mail that included the study’s consent form. For this study, we approached attending physicians working both at UNM Hospital and the New Mexico Veterans Affairs Health Care System since both organizations train fellows in our program. We also approached all 1st- and 2nd-year fellows, except for the two fellows that were study co-investigators. There was a 100% response rate for all surveys sent form the consented study participants.

2.4. Survey

The survey was created by the research group and consisted of two sections, one which collected respondent demographic information and a second focused on the evaluation of each of the proposed EPAs. The emailed survey included a training tool with the scope to explain EPAs and the entrustment decision-making process and to provide participants with instructions on how to complete the survey. The survey respondents were asked to evaluate each proposed EPA on a 5-point Likert scale, ranging from 1 should definitively be excluded to 5 should definitively be included.

After each proposed EPA, the respondent could add comments or items that were reviewed for the next cycle/year of the study (implementation and evaluation cycle). Three rounds of survey administration occurred between 18 April 2024 and 28 May 2025.

2.5. Consensus Standard

A discussion among the study investigators took place after each response cycle to finalize which EPAs would be included in the following cycle of refinement. Consensus was achieved when the mean rating was ≥4 with a standard deviation of <1 and at least 75% of respondents rated the EPA ≥ 4 [8]. A total of three cycles were performed to reach a final consensus.

The recruited faculty members and the fellow were, as expected, familiar with the professional work required of nephrology fellows under training, but they did not necessarily have a sufficient understanding of what EPAs are. To familiarize the study participants with the concept and necessity of developing EPAs, we provided them with a training PPT before their involvement in the study. Between study participants, the main point of disagreement was on whether all the EPAS that were related to kidney transplantation training were necessary or not. At the end of the 3rd round, 3 transplant-related EPAs were eliminated, and we concluded that the remaining 2 transplant EPAs are effectively “what everybody is doing” no matter whether they are a general or a transplant nephrologist.

A discussion on the suggestions and comments from the study participants took place among the research group members after the completion of all 3 rounds. Many comments from the co-investigators were about the way we deliver the EPAs and the number of observations that are needed to evaluate each EPA rather than their actual content. Regarding the ability to deliver to patients the news of the need to initiate dialysis, the group decided that we could use standardized patients focusing on dialysis decision-making. Social determinants of health were another point of discussion since many patients are dealing with medication co-pays and limited access to transportation, and there is a significantly higher incidence of cardiovascular events and mortality compared to those with financial stability [11]. Finally, most study participants and the research team agreed that the fellows need greater exposure to chronic dialysis modality patients, including home dialysis.

2.6. Matching the EPAs to the Published ACGME Milestones

Two independent reviewers (HS and MER) mapped our final list of EPAs onto the ACGME Program Requirements for Graduate Medical Education in Nephrology. Mapping the EPAs onto the ACGME requirements allowed us to explore whether the final list of EPAs successfully represented the overall scope of nationally required nephrology training. All developed EPAs were matched to the published ACGME guidelines.

3. Results

3.1. Survey Respondents

Our study population consisted of nine survey respondents: eight faculty nephrologists recruited from both the UNMH and the VA hospital and one nephrology fellow. Initial invitation to participate in the study panel was sent to 28 total individuals (20 nephrology attendings and 8 nephrology fellows). Five of the invited nephrology attendings and two of the invited fellows (study co-investigators) were involved in the development of the initial set of EPAs. All attendings and the fellow that initially consented to participate in the study completed all three circles of the survey. The collected demographic data for the participant group are displayed in

Table 1. Participant group demographics.

Parameters
Age (mean ± SD)	47.2 (±8.7)
Females (%)	44.4
Attendings (%)	88.9
Fellows (%)	11.1
Years of clinical experience (median, IQR) a	12.5 (17.5)

^a Years of clinical experience included both academic and private practice environments.

.

3.2. Creating the EP List and Building Consensus

The initial EPA creation resulted in 25 separate EPAs. The consensus-building process led to the removal of EPAs numbered 11, 13, and 14 (Table S2). A final list of 22 EPAs, including the marked removed EPAs, is displayed in

Table 2. Final list of the developed EPAs.

Final EPA List
Chronic Kidney Disease
(1) Assessing and providing an initial plan for investigation/work-up and management of patients with CKD (reduced eGFR and/or proteinuria).
(2) Monitoring and providing medical management and comprehensive care for patients with stable kidney disease or progressive kidney dysfunction.
(3) Facilitating patients’ transition to an ESKD treatment modality or to end-of-life care and being able to identify eligible patients and discuss all available therapeutic options including, if applicable, the provision of kidney supportive measures with the patient and/or their caregivers.
Hemodialysis
(4) Ordering and adjusting dialysis prescriptions for uncomplicated patients with ESKD.
(5) Assessing and providing initial management for patients with common (a) acute (e.g., hypotension, bleeding, cramping, nausea/vomiting, etc.) and (b) chronic complications of HD (e.g., itching) and HD access (e.g., thrombosis, stenosis, infections, etc.)
(6) Providing longitudinal management for patients receiving chronic HD.
Peritoneal Dialysis
(7) Ordering and adjusting PD prescriptions for uncomplicated patients with ESKD.
(8) Assessing and managing the care of patients with (a) infectious and (b) non-infectious complications of PD.
(9) Providing longitudinal management for patients receiving chronic PD.
Kidney Transplantation
(10) Assessing and providing initial management for transplant patients with acute kidney allograft dysfunction.
(12) Providing postoperative care for kidney transplant recipients with an uncomplicated or complex course.
(15) Monitoring patients receiving immune-modulating therapy and managing kidney transplant-related complications.
Acute Kidney Injury
(16) Assessing, establishing, and providing an initial comprehensive management plan for patients with AKI.
(17) Ordering and adjusting RRT prescriptions for patients with AKI and other acute/urgent indications for RRT.
Electrolyte and Acid–Base Abnormalities
(18) Assessing and providing an initial investigation and management plan for patients with complex fluid and electrolyte abnormalities.
Hypertension
(19) Management of uncontrolled hypertension in hospitalized patients with CKD assessing their complications and target of treatment.
(20) Assessing and treating patients with difficult-to-control idiopathic or suspected secondary hypertension.
Kidney Stones
(21) Initial and longitudinal management plan of patients with kidney stones.
Glomerular Diseases
(22) Initial diagnostic and management plan for patients with nephrotic or nephritic syndrome.
(23) Longitudinal management of GN patients receiving immune or non-immune therapy and management of therapy-related complications.
Pregnancy
(24) Integrating knowledge of the effects of pregnancy, pregnancy outcomes, kidney disease, and its treatment in the care of women with kidney disease.
Teaching and Education
(25) Delivering scholarly teaching to a variety of audiences, including peers, junior trainees, and/or other health professionals.
Removed
(11) Admitting patients to undergo kidney transplantation. a
(13) Assessing and providing initial management for patients with common infectious and non-infectious complications of kidney transplantation. a
(14) Assessing the eligibility of patients with kidney disease for kidney transplantation. a

Abbreviations: CKD, chronic kidney disease; eGFR, estimated glomerular filtration rate; ESKD, end-stage kidney disease; HD, hemodialysis; PD, peritoneal dialysis; AKI, acute kidney injury; RRT, renal replacement therapy; GN, glomerulonephritis. ^a indicated that EPAs were initially part of the category “Kidney Transplantation“ before removal.

.

Each finalized EPA achieved strong agreement across participants with a mean rating of ≥4.4 (SD ± 0.34). The average ratings given on EPAs during each round by the study panelists stratified by years spent in service into categories of <10 years, 10 to 20 years, >20 years, and fellowship are shown in Supplemental Table S3.

3.3. Matching the EPAs to the ACGME Milestones

The final list of developed EPAs was matched to 38 different ACGME sub-competencies outlined within the program accreditation guidelines by two independent study members (HS and MER), and consensus was reached (

Table 3. Final EPA list matched with ACGME sub-competencies.

EPA Number a	ACGME Sub-Competencies b
1	IV.B.1.b).(1).(a).(ii)
	IV.B.1.b).(2).(a).(i)
	IV.B.1.b).(2).(a).(ii)
	IV.B.1.b).(2).(b).(iv)
	IV.B.1.c).(1).(j)
	IV.B.1.e)
2	IV.B.1.b).(1).(a).(ii)
	IV.B.1.b).(2).(a).(ii)
	IV.B.1.c).(1).(a)
	IV.B.1.c).(1).(g)
3	IV.B.1.b).(1).(a).(vi)
	IV.B.1.b).(2).(a).(iii)
4	IV.B.1.b).(1).(a).(vi)
	IV.B.1.b).(2).(a).(iii)
	IV.B.1.b).(2).(b).(i)
	IV.B.1.b).(2).(b).(iii)
	IV.B.1.c).(1).(a)
	IV.B.1.c).(1).(b)
	IV.B.1.c).(1).(c)
	IV.B.1.c).(1).(o)
	IV.B.1.c).(1).(q)
5	IV.B.1.b).(2).(b).(i)
	IV.B.1.c).(1).(b).(v)
	IV.B.1.c).(1).(b).(vi)
6	IV.B.1.b).(2).(a).(iii)
	IV.B.1.b).(2).(b).(i)
	IV.B.1.b).(2).(b).(iii)
	IV.B.1.c).(1).(b)
	IV.B.1.c).(1).(o)
	IV.B.1.c).(1).(r)
	IV.B.1.f)
7	IV.B.1.b).(1).(a).(vi)
	IV.B.1.b).(2).(a).(iii)
	IV.B.1.b).(2).(b).(iii)
	IV.B.1.c).(1).(b)
	IV.B.1.c).(1).(c)
	IV.B.1.c).(1).(o)
	IV.B.1.c).(1).(q)
8	IV.B.1.b).(2).(b).(iii)
	IV.B.1.c).(1).(b)
	IV.B.1.c).(1).(q)
	IV.B.1.c).(1).(r)
9	IV.B.1.b).(2).(a).(iii)
	IV.B.1.b).(2).(b).(iii)
	IV.B.1.c).(1).(b)
	IV.B.1.c).(1).(r)
	IV.B.1.f)
10	IV.B.1.b).(1).(b)
	IV.B.1.c).(1).(l).(vii)
12	IV.B.1.b).(1).(b)
	IV.B.1.c).(1).(l)
15	IV.B.1.b).(1).(b)
	IV.B.1.c).(1).(a)
	IV.B.1.c).(1).(h)
	IV.B.1.c).(1).(l)
16	IV.B.1.b).(1).(a).(i)
17	IV.B.1.b).(1).(a).(i)
	IV.B.1.b).(2).(b).(ii)
	IV.B.1.c).(1).(a)
	IV.C.5.h)
18	IV.B.1.b).(1).(a).(iii)
	IV.B.1.c).(1).(e)
19	IV.B.1.b).(1).(a).(ii)
	IV.B.1.b).(1).(a).(x)
	IV.B.1.c).(1).(d)
20	IV.B.1.b).(1).(a).(x)
	IV.B.1.c).(1).(d)
	IV.B.1.f)
21	IV.B.1.b).(1).(a).(iv)
	IV.B.1.c).(1).(f)
	IV.B.1.c).(1).(i)
22	IV.B.1.b).(1).(a).(ix)
	IV.B.1.b).(2).(b).(iv)
	IV.B.1.c).(1).(g)
	IV.B.1.c).(1).(h)
23	IV.B.1.b).(1).(a).(ix)
	IV.B.1.c).(1).(a)
	IV.B.1.c).(1).(h)
24	IV.B.1.b).(1).(a).(xi)
25	IV.D.1.b)
	IV.D.2.b)

^a EPAs numbered 11, 13, and 14 were not matched due to elimination during consensus-building process. ^b For details regarding numbered competency please refer to ACGME Program Requirements for Graduate Medical Education in Nephrology 2024 [12].

). Throughout the matching process, sub-competencies IV.B.1.c).(1).(a), IV.B.1.b).(2).(a).(iii), IV.B.1.b).(2).(b).(iii), and IV.B.1.c).(1).(b) were matched the most, with each being matched to five separate EPAs.

4. Discussion

4.1. The State of Competency-Based Medical Education

The ACGME has established a structured competency-based framework for the assessment of trainees in all ACGME-accredited programs, including adult nephrology fellowship [13,14]. While the ACGME has created these frameworks to evaluate fellows during training, the actual competencies remain broad and abstract in their language, making them challenging to translate into use in practical evaluation. For example, within the patient care domain, the sub-competency Chronic Dialysis Therapy (PC2) outlines progressive levels of proficiency [12]. At the highest level, the milestone describes a fellow’s ability to “[identify] the complexities of providing quality care to a population of patients receiving dialysis” and “[anticipate] and [manage] the breadth of comorbid medical and technical complications in the patient on dialysis, including when dialysis is not appropriate” [13]. While this may seemingly encompass in a broad sense the necessary skills a nephrology fellow needs to be proficient in the care of patients on chronic dialysis, it lacks specificity on how these skills can and should be demonstrated, observed, or assessed in routine clinical practice. This ultimately creates a gap between regulatory language and the day-to-day supervisory judgments that a program director and supervising physician must make when determining a fellow’s readiness in managing this aspect of patient care. Our study sought to address these limitations by using the framework of EPAs to create observable and trackable units of clinical work that reflect the most essential responsibilities a practicing nephrologist must be able to perform independently regardless of their site of practice at the completion of their fellowship.

Although CBME has gained increasing traction in postgraduate training across multiple specialties, the implementation of structured assessment tools in the field of nephrology remains limited and highly variable [15]. Several efforts over the past decade have made strides in addressing this gap by providing foundational work in milestone development, curricular mapping, and EPA development [5,16,17,18]. One of the earliest calls for nephrology-centered competency structures came from a proposal published by Yuan et al. in 2013 [19]. They advocated for milestones that are both quantitatively measurable and objectively defined to reduce both evaluator bias and improve reproducibility among programs [19]. They also emphasized the value of simplicity and parsimony in evaluation metrics, meaning that a higher number of milestones would not equate to better outcomes. Rather, simpler and succinctly defined milestones would lead to better tracking, evaluation, and outcomes for the program [19].

4.2. Development of EPAs in the Field of Nephrology

Our EPA development efforts directly echo several recommendations made by Yuan et al. [20], such as their call for simplicity and objectively defined metrics. For example, if we recall the ACGME nephrology milestone PC2, our equivalent EPAs that cover the patient care aspect of chronic dialysis are addressed by EPAs 4, 5, and 6 for hemodialysis and 7, 8, and 9 for peritoneal dialysis. While the ACGME milestone uses a generalized statement to cover all aspects of chronic dialysis treatment, our six EPAs each descriptively call for evaluation of a specific aspect of chronic dialysis care. EPA 4, for example, addresses ordering and managing prescriptions of patients on hemodialysis, while EPA 5 focuses on managing the complications of hemodialysis. These EPAs can be more readily linked to skill performance.

Subsequent work by the same authors further explores the operationalization of nephrology-centered milestones and provides a framework for mapping ACGME’s initial 23 IM sub-competency milestones to a nephrology schema [20]. They describe a limited set of five EPAs and associate each with suggested tools for assessment. However, two main limitations apply to this initial framework. The study primarily focused on curricular mapping and descriptive categorization of nephrology-based competencies. Secondly, the developed EPAs were not formulated through a structured consensus process and limited in scope, addressing only five aspects of nephrology practice. In contrast, our study centered around the goal of developing a widely generalizable set of relatively comprehensive nephrology-specific EPAs that were derived from expert agreement and aggregated through strict consensus criteria. Furthermore, our study began with principle clinical proficiencies and worked to match each to preexisting ACGME core competencies and educational program requirements. This approach grounded our EPA list not only in educational theory but in practice-based consensus.

Outside the U.S., efforts to redefine nephrology training assessment with EPA-based models have also emerged. In Canada, the Royal College of Physicians and Surgeons launched the Competence by Design (CBD) initiative to transition postgraduate training from a time-based to an outcome-based structure [21]. As a part of this national reform, nephrology-specific EPAs were developed for both adult and pediatric settings. These EPAs were categorized across four stages of training and were each mapped to CanMEDS milestone domains [21]. Our study parallels this work but leans on ACGME competency standards to be widely applicable in USA postgraduate training. Another point of distinction is that, while the Canadian framework was led by a central regulatory body with national authority for implementation, our EPA development reflects a grassroots consensus from USA-based nephrology educators and program leaders working at a fellowship that often trains fellows destined for both academic and community-based careers. Our study highlights an important first step in aligning national fellowship expectations in nephrology with an EPA-based framework grounded in USA-based practice realities.

We chose to utilize the ACGME Program Requirements for Graduate Medical Education in Nephrology 2024 as the primary resource for milestones for matching our created EPAs rather than the ACGME’s Nephrology Subspecialty Milestones [12,13]. While the subspecialty milestones serve as critical developmental benchmarks for assessing fellow progression, they continue to remain broad and difficult to translate into discrete clinical tasks. The accreditation requirements provide more structured and detailed sets of competencies for various domains of nephrology training that can be matched to our EPAs. By grounding our EPA framework in program accreditation standards, we ensure that our EPAs not only perform their function as observable and assessable units of training but also align with national expectations for fellowship curricula.

4.3. Strengths and Limitations

This study offers several notable strengths. First, the EPA development process was grounded in consensus from an expert panel that included both the nephrology faculty and nephrology fellows. This ensured that the resulting EPAs reflected not only the supervisory expectations of the faculty but also the lived educational realities of trainees. Previous work has highlighted that there is potential disconnect that can occur between the perception and adoption of EPAs and their benefits from the viewpoint of faculty and trainees [22]. While our study only had one fellow, we believe that due to the small study size in total, the fellow would have had appropriate ability to carry out any concerns during the EPA developmental process. Second, participants were recruited from both UNMH and the affiliated Veteran Affairs Medical Center, allowing us to integrate insights from nephrologists into two distinct practice settings serving distinct patient populations.

However, the study is not without its limitations. Most significantly, despite our goal to make it widely applicable, the EPA development process was conducted at a single fellowship program which could limit the generalizability of the EPA framework to other programs across the U.S. Future multicenter studies should aim to validate these EPAs nationally to build consensus on the adaptability of these developed EPAs. Another significant methodological limitation for our study was the number of study participants involved. Although the Delphi methodology is well suited for achieving structured expert consensus, the relatively small panel size of nine may have limited the breadth of perspectives represented. A smaller panel may increase susceptibility to individual bias and may reduce the generalizability of the consensus findings. While Delphi panels commonly include small groups of subject-matter experts and studies have shown that their strengths may lie in that flexibility, our results should nevertheless be interpreted within the context of this sample size [8,21,22,23,24]. Future studies including a larger and more diverse group of panelists, including individuals across multiple institutions and those who were not present in the initial EPA creation cohort, may help strengthen the robustness and external validity of nephrology EPA creation.

Moreover, we are cognizant of various conceptual and practical pitfalls that come with the implementation of competency-based education frameworks. One concern is that the implementation of EPAs may increase documentation burden for both faculty and learners. Previous studies have shown that both fellowship programs, at least in pediatrics, using and not using EPAs indicate their major barrier for implementation as work burden [5,23,25]. Indeed, the implementation of an EPA framework comes with additional logistical burden on programs which individual institutions will need to account for. Our current study does not address the infrastructure or resources required to implement EPA-based assessment frameworks, but this should be the subject of future research. Recognizing these limitations underscores the importance of future research aimed at refining these EPAs and determining optimal approaches to EPA implementation. The scarcity of nationwide literature for nephrology EPA implementation in the United States becomes the crux of the purpose of this study. The hope is that the development of these EPAs, specifically in the field of nephrology, will add both to the increasing implementation of EPAs throughout postgraduate medical education and allow the greater medical education system to benefit from studying trainee outcomes with CBME formats.

5. Conclusions

This study represents a first step in establishing a structured, consensus-driven EPA framework tailored to nephrology fellowship training in the USA. By aligning each EPA with ACGME accreditation requirements, milestones, and core competencies, we aimed to develop a set of observable and assessable tasks that encompass the most crucial aspects of nephrology fellowship training. The finalized set of EPAs developed will become a foundation upon which standardized assessment, curriculum development, and program evaluation will be built. The subsequent phase of this study will utilize these developed EPAs and study their practical application and potential barriers such as documentation burden, faculty training, and integration with existing tools for implementation in assessing year 1 and year 2 nephrology fellows at the UNM Nephrology Fellowship Program. Multicenter validation in the future will be vital for both corroboration of our developmental method for EPAs in nephrology and to also mitigate any single study and small sample limitations of this study.